Open AndreyRGW opened 5 days ago
The method used for CFG++ doesn't work in A1111, last_noise_uncond is set differently. I've done some testing to see if I can easily make it work, without success. I'll probably just add a check to not show the samplers in A1111.
NansException: A tensor with NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check.